should be at least one target in each quadrant. The scanner’s
maximum range and accuracy may limit effective scan coverage.
Figure 1. Target setup layout A flat target can be stuck on a surface as shown in Figure 2a, a
spherical target shown in Figure 2b and a tilt-and-turn planar target shown in Figure 2c can be mounted to a tripod. Different
laser scanning targets are designed for different specific distances; they should be scanned within the distance in order to
get enough density to pinpoint their centers for registration purposes. A good target should have enough point cloud
coverage and the center of the target should be clear as shown in Figure 3a. If the distance from the scanner to the target exceeds
the manufacturer’s recommended distance, it is hard to pinpoint the center of the target as shown in Figure 3b increasing error.
a. Sticky paper
target b. Sphere
target c. Tilt and
turn planar
target Figure 2. Different types of targets
a. Within the recommended
distance, with enough point
cloud coverage. b. Beyond
recommended distance, with
very low point cloud density.
Figure 3. Different distanced between scanner and target
Capture of point cloud data For documenting as-built conditions, the density of point clouds
needs to be high enough to produce CAD drawings, a 3D mesh model, and panoramic point cloud pictures. A high resolution
scanner is needed in underground construction such as tunnel projects. Also, given the tough and confined environment of a
tunnel project, it is preferable to use a fast-time scanning equipment. A phase-based scanner is recommended for the
documentation of a tunnel construction. The scanning geometry plays an important role in the quality of
the resulting point cloud. Depending on the purpose of the project and the type of scanner available, the scanner should be
set up to have enough coverage in the given environment to achieve high accuracy in the point cloud data. Due to the
conditions in a construction environment, an ideal setup is not always possible. The location of each scan depends on the field
situation, but should have as much unobstructed coverage as possible.
Processing of the point cloud data Registration is the process of integrating several scans into a
single coordinate system. The registration process performs optimal alignment transformations to make sure that the targets
used as constraints are aligned as closely as possible. After registration, the coordinates of all point cloud data are the same
as the control system. A full 3D mesh model is created to offer a better visualization of the as-built tunnel as well as sets of
panoramic point cloud pictures. Mapping the panoramic point cloud pictures on the web-
based service. A panoramic image of the point cloud data is created for web-
based sharing and viewing. To be distinguished from the panoramic image, the point clouds have accurate location
information that can be measured, marked, and hot linked for sharing with different project participants. The as-built
information has great value for the client and can be stored for future use. Each scan generates a set of panoramic point cloud
pictures. The location information is stored in an XML file. By extracting the information in the XML files from each scan,
the sets of pictures can be mapped to an online map service, such as Google Map. In this case, point clouds can be accessed
anytime easily without limitation. The final deliverable is URL containing a single Google map with sets of panoramic point
cloud pictures on it. If the data are considered to be sensitive and should not be available to the general public, invitation
codes can be created for specific users.
4. CASE STUDY
The Chicago Transit Authoritys CTA Red Line is Chicagos busiest rail line with an average weekday ridership exceeding
250,000. Its a vital artery of the city, and runs 24 hours a day, 365 days a year
—ideally. Chicagos Dynasty Group were the projects lead surveyors, and conducted the project with the
most sophisticated corridor survey in the Authoritys history, a massive logistical effort that ultimately included traditional, ,
laser, and ground penetrating radar surveys of the entire route, plus a geo-referenced video. This study is based on the portion
of this project that documents and visualizes the tunnel portion of the CTA Red Line.
Pre-scanning phase. The CTA Red Line tunnel stretches from the south portal
around 18th St. and S Wentworth Ave. to the north portal close to the Armitage Station with a total length about 4 miles
as shown in Figure 4. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences,
Volume XL-2W2, ISPRS 8th 3DGeoInfo Conference WG II2 Workshop, 27 – 29 November 2013, Istanbul, Turkey
This contribution has been peer-reviewed. The peer-review was conducted on the basis of the abstract. 141
Figure 4. CTA Red line tunnel It contains two tracks going north and south. The tunnel passes
through downtown Chicago and under the Chicago River with a variety of geotechnical characteristics. Because the Red Line
runs 24 hours a day, as-built documentation presented a big challenge. In order to minimize the noise created in the data by
the passing trains, scans were performed during the night when traffic is low. In addition, flaggers helped to stop the train while
scanning. To have maximum coverage of the tunnel, 500 scans were performed in a period of two months. Ground surface
networks were defined and stretched out to provide control in the tunnel. Only tilt and turn targets were used in the tunnel
environment. The targets were set up on tripods as shown in Figure 5a or attached to the steel columns on platforms using a
magnetic clip as shown in Figure 5b.
a. Target on tripod b. Target attached
to column Figure 5. Location of targets
Capture of point cloud data The Leica HDS 6200 scanner was selected to perform the scans
in the tunnel. This system is a phase based scanner offering the fastest scan rates up to 1 million points per second for high-
accuracy in as-built surveys. It is especially suitable for the as- built tunnel documentation because it operates in a very short
time window for capturing high density, high definition point cloud data to avoid the occlusions caused by the passing trains.
The scanner was typically located at the edge of the tunnel or on a platform, to have maximum coverage of the tunnel as shown
in Figure 6. Figure 6. Location of the scanner
In order to acquire the most accurate results from the scanner, the regular interval in between the two scans was set up to be
less than 20 m to have at most 5 mm accuracy. A super high resolution was selected to give roughly 7 minutes per scan.
More than 500 scans were performed during the two-month period to cover the entire tunnel as shown in Figure 7.
Figure 7. Point cloud of the tunnel
Processing of the laser scan point cloud After performing a 3D laser scan, the raw point cloud data need
to be processed. Cyclone was selected as the software to process point cloud data because it is a software developed by Leica, the
manufacturer of the laser scanner used in the study, hence avoiding compatibility problems in between the software and
the equipment, and avoiding the cost of purchasing a new software since Cyclone comes with the Leica laser scanner.
Cyclone supports the use of targets to geo-reference scan data, as well as the ability to align overlapping areas of point clouds.
More than five hundreds scans were registered together in the State Plane coordinate system to generate one complete model
space to perform survey documentation as shown in Figure 8a, to establish a 3D mesh model as shown in Figure 8b, and to
create hundred sets of panoramic point cloud pictures.
a. Survey documentation b. 3D model
Figure 8. Documentation of the tunnel International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences,
Volume XL-2W2, ISPRS 8th 3DGeoInfo Conference WG II2 Workshop, 27 – 29 November 2013, Istanbul, Turkey
This contribution has been peer-reviewed. The peer-review was conducted on the basis of the abstract. 142
Using uGridd to map Truview to Google Map. One of the products created by point cloud data is called
“Truview”. Truview is a set of panoramic pictures composed of point cloud data. The engineer can extract real 3D coordinates
and accurately measure distances from the pictures. Also, hotlink and markups can be created and shared with other
people for better communications as shown in Figure 9.
Figure 9. Panoramic point cloud picture The problems in this project were that 1 sections of the tunnel
looked very similar to each other, and 2 there were no references to help to locate an object through the entire five
mile length of the tunnel. It was difficult and time consuming to reference the as-built data to the right location. This problem
was solved by geo-referencing the as-built data to a web mapping system.
uGRIDD is a web-based service provider that offers geo- referencing of infrastructure data. One of
uGRIDD’s functions called “Truview2Map” was used in this research to geo-
reference hundreds of truviews to Google map. By extracting the information in the XML file in the truview folder and
projecting it to the State Plane coordinate system and to the WGS 84 coordinate system, the location of each truview was
located on Google map as shown in Figure 10.
Figure 10. Truviews on Google Map Each triangle represents a scan setup. When one clicks on a
yellow triangle, detailed information, such as scan name, scanner model, number of point cloud, and scan date show up in
a bubble as shown in Figure 11.
Figure 11. Truview’s information Some data may be sensitive and perhaps should not be available
to the public. For security reasons, detailed information about the CTA tunnels should not be published without any
restrictions. Invitation codes were created to secure the information. Only authorized persons who have been issued
invitation codes could view the information. The final product was a simple URL that could be opened and shared with other
authorized project participants.
5. CONCLUSION
